Seminar: Chenwei Xiao

Humans manage the majority of the land surface but the consequences of management on ecosystem resilience is still an open question. This study aims to understand the land cover and land management effects on ecosystem resistance to drought stress and temperature. We used satellite observations of greenness (kNDVI) and a local moving window method to investigate the local effects of land cover changes on ecosystem resistance. We used dynamic global vegetation models (DGVMs) to disentangle the effects of land cover changes from those of climate change and elevated CO2. Our results indicate that ecosystem resistance to drought decreases when forests are converted to grasslands or croplands, consistent with previous studies using satellite-based estimates of vegetation optical depth (VOD), or empirically upscaled estimates of GPP. DGVMs simulate similar resistance patterns with kNDVI using vegetation biomass. Models also predict substantial historical land cover change effects on ecosystem resistance. However, the substantial effects of the forests-croplands transition on ecosystem resistance as suggested by the analysis of the kNDVI data are not simulated. Our study demonstrates large variability among models and the need to improve vegetation resilience representation.